AI and the Future of Financial Regulation

Insights from the Inaugural Finance AI Research Lab (FAB) Symposium

Barry Quinn CStat
Co-Director of FAB

FAB launch

From right to left: Barry Quinn (QBS), Chris Clements (FinTrU), Adrian Johnston (Catalyst Inc), Pavle Avramovic (FCA),Carla McGlynn (Citi), Fearghal Kearney (QBS), Daniel Broby (UU)
Figure 1: Companies in attendance

Keynote Highlights - FCA’s Approach

  • Title: Emerging Technologies & Regulation by Pavle Avramovic
    • Importance of understanding new technologies for regulation
    • Deep dive into synthetic data and Privacy Enhancing Technologies (PETs)
    • Horizon Scanning Framework and the challenges of Quantum Computing and DeFi

Academic Insights

  • Title: AI and Compliance Audit by Professor Hui Wang
  • Bullets:
    • Advances in NLP for auditing complex regulations
    • Machine learning algorithms for detecting anomalies
    • The role of Computer Vision in ensuring compliance
  • Title: Financial Auditing in the Blockchain Era By Daniel Broby
  • Bullets:
    • Challenges of transaction malleability, DAOs, and blockchain forks
    • The role of auditors in blockchain custody and cross-chain transactions
    • Insights into time-locked and multi-signature transactions

Panel Discussion: Opportunities & Risks

  • Title: AI in Financial Services
  • Bullets:
    • Opportunities for Northern Ireland in AI and Financial Services
    • Fintech challenge of manage the risk of employee displacement due to AI adoption
    • More bespoke skills training on domain-specific AI application
    • NI as a AI digital sandbox which spans two regulatory jurisdications

Current FAB projects

  • Market Manipulation Theory and AI
  • The Economic Cost of Cultural Displays in NI
    • Mapping the Longitudinal Patterns of Cultural Displays
  • Computer Vision and Car Insurance Pricing (Liberty IT)
  • Fund management Regulations, Rules-based Engines, and the use of LLMs (Funds Axis Ltd)

Early FAB Output

  • Title: Evolutionary Multi-Objective Optimisation for Large-Scale Portfolio Selection
  • Authors: Weilong Liu, Yong Zhang, Kailong Liu, Barry Quinn, Xingyu Yang, Qiao Peng
  • Affiliations: Guangdong University of Technology, Shandong University, Queen’s University Belfast; FAB
  • Publication: IEEE Transactions on Evolutionary Computation
  • Objective: Present a cutting-edge model for optimising large-scale investment portfolios, integrating traditional and novel securities.

Early FAB Output

Innovative Model & Algorithm

  • Approach: The study introduces a hybrid model that adeptly handles both established and newly listed securities, employing mean, variance, and skewness of returns. It uniquely combines random and uncertain variables to reflect real-world complexities.

  • LSWOEA Algorithm: A novel evolutionary algorithm optimised for large-scale portfolios, proving more effective than existing strategies in testing.

Early FAB Output

Contributions & Applications

  • Academic Impact: Significantly advances computational finance by offering a robust framework for managing large portfolios under uncertainty.
  • Practical Value: Provides institutional investors with a sophisticated tool for improving portfolio diversification and risk-adjusted returns, bridging the gap between theoretical research and real-world financial practices.